Error due to 'pre no' in run command

LAMMPS
, ,
1

LAMMPS ver Jun 2023 and Nov 2022

In the code below I get a segmentation fault error in this code if I add ‘pre no’ to run command,

‘pre no’ cause another error in a another script which is
Dump 1 every variable logsteps returned a bad timestep: 0 (src/output.cpp:509)
in the other code where I am using logfreq3, I get the error. Although, when I print the current timestep it gives 0 and printing the variable it gives 1. So, the variable is greater than the timestep, but still gives an error.

dimension       2

units           lj
atom_style      atomic
boundary        p p p
region          box block 0 10 0 10 -0.2 0.2
create_box      1 box
create_atoms    1 random 1 123456 box
mass            * 1.0


run 0

dump            1 all custom 1 data.restart_small.xyz id x y vx vy

run 1000 post no pre no
`
``
2

Please see the documentation. You may use pre no only when you make no modifications to your simulation. Adding a dump is a modification, so using pre no will skip necessary initialization and thus leads to the segfault. You must use pre yes in such cases. This is not a bug, but incorrect input.

That is a different code with a different error message. Impossible to debug this way.